Brian Ratliff

At higher yaw angles, the drag will very likely be higher with a narrower rim of the same height.

Skinny wings have low stall angles. Thick wings have higher stall angles, especially when they are thickest at the midpoint of the wing rather than at the leading edge, at the expense of (slightly) higher zero angle drag. It also means that a thicker rim will be easier to control in cross winds, and it's an admission that minimizing zero angle drag isn't the be-all and end-all of the equation. Drag at high yaw angles, as well as a high stall angles to help with controllability in crosswinds, are important parts of the equation as well.

Hida Yanra

well, being as how I just picked up a hed wheel - who makes these pad holders?

tanhalt

Yup, as the guys at VN apparently discovered when they did their testing of the Stinger 9s.

And just recently I was told that the 2011 models will have a shallower tire bed to allow narrower than 23C tires...although there was also NO details on what effects that has on the drag curves they show...

pjcampbell

right this means on a zipp clincher you have to run a 20mm clincher basically yes? i ran these and was very happy with them to go straight, but i am looking for an all around training, racing wheel to ride on crappy roads.

Racer Ex 

You'd certainly think so. Build a 17mm front that saves 15w and you'd be a rich guy.

Until then we're stuck with what works best off paper.

Racer Ex 

The more time I spend in the tunnel and testing stuff, the more I've come to realize how complex wheel/tire design is. You have a fair number of mechanical constraints to deal with, and the aerodynamics of something that has 360 degree wind exposure (I'm a leading edge, no wait, I'm a trailing edge), moving spokes, and a hub sitting in between two blades that are also directing air...that's not just plugging in a formula and presto.

And that's just looking at the front and not addressing control. I ended up going to the less slippery zero yaw rear because taking the other out to a certain yaw angle it fell off the edge of the earth. And then you better watch what happens after that air leaves and starts hitting other parts on the bike; I could tell you some stories about modeling clay finding 3 watts on one wheel set and having zero effect on another.

Bikes constrained by UCI parameters are really complex animals.

mzeffex 

I have a set up 23 mm wide tubular rims I'm racing CX on now. I plan to put road tubies on there for racing crits in Spring. Can I put a 22 mm tire on there like so many are (conti, for one) or do I need to stick to a 23 such as the S-works?

waterrockets 

The rotation thing is hard for me to think about in aerodynamic terms. It's even worse than spinning around the axle, because the bottom of the tire isn't moving at all, and instaneously, everything is actually spinning around the bottom of the wheel and not the hub. Then the top of the tire is traveling 2x fast as the hub, but it's a different angle than the parts traveling at the same speed as the hub....

Yeah, so as you say, you pretty much have to test it and work hunches and test again. Sounds fun, and I hope I can participate one day. It would be cool to build a rig that spins a wheel around a circumferential point rather than the hub to test drag -- but that wouldn't account for the fork -- so you really need a tunnel or coast-down.

Brian Ratliff

1+ Extremely complex, which is why I hedge my comment and speak only in terms of wing profiles, not rim profiles. Very difficult to model without those fancy computers over at Sandia they use to model nuclear blasts.

Just out of curiosity, what is the typical uncertainty for wind tunnel measurements? I assume the force meter doesn't just peg out at a precise value for five minutes as the reading is taking place... +-5%? 10%?

Brian Ratliff

It's actually simpler than you are making it out to be. Vectors are additive... from the perspective of the air, the wheel is spinning around it's hub. That's the perspective that matters here; we are worried about how the bike moves though the air, or conversely, how the air moves around the bike (one or the other, air or bike, is fixed). From the perspective of a fixed point on the ground, the wheel is rotating around the instantaneous ground contact point... however, if you want to do any calculations involving the air moving around the bike, you have to add in the wind speed vector to every other velocity vector, which, once you work all the vector additions out, puts everything back to the rim spinning around the hub.

carpediemracing 

I ran an Evo CX not knowing the 23mm thing all year. 21mm I think. I have some 23mm Bontrager tubulars I'll be gluing sometime this winter. Curious what the difference will be.

tanhalt

What's wrong with 19mm?...there's an awfully good tire out there still (for a short while, anyways) that works REALLY well on rims designed for that width of tire (think DuPont/Specialized/H3 trispoke).

Heck, just show me the "off paper" results that a narrow Jet 90 with a Bontrager Aerowing TT 19C tire is slower than any tire you choose on a C2 Jet 90.

tanhalt

Actually...fairly good CFD software (including rotating elements) is available currently that runs on PCs. No "supercomputers" needed ;-)

tanhalt

I'm talking about width at the brake track, not the overall rim width...plus, you need to look at some of the stuff Zipp has done lately with the Firecrest models where they majority of the "gains" were reaped by looking at how to make the section better when the tire is the trailing edge. Think of the wind hitting the rim section from below in the following pic


https://forum.slowtwitch.com/gforum.c...ring=;#3054556

Racer Ex 

I'm sure it will vary from tunnel to tunnel but in relative terms it appears pretty small. Our runs with the test dummy (me) pedaling were within 1-2%, the bike alone numbers were well below that. That's for A2, the other tunnel I used was slightly less consistent, but only slightly. And granted I'm a pretty good test dummy if the picture overlay is any indicator.

In absolute terms between tunnels, I don't know and because I went into each tunnel on entirely different rigs, I don't have a basis for any anecdotal comparison.

Racer Ex 

Nothing wrong with a 19mm tire, I used the 17mm as an example of how function would trump whatever number showed up on the tire/rim.

Didn't run that exact combo but I did run several "narrow" rims with matching tires, all tubulars. C2 beat all of them. At all the yaw angles we tested with the wheels spinning.

If you have tunnel data for the old Jet with that tire compared to the C2 Stinger or C2 Jet, I'm all ears. Or if you want to send it out with me to the tunnel this winter, I'd go ahead and run it. Albeit it would be in my frame which can influence the number some. I'm not wedded to any particular dogma or trying to prove any theory, I just go with whatever produces the best numbers.

Saw Al's Crr number for that tire BTW...that combo would have to be a fair bit better to move me off of what I have.

waterrockets 

That will account for wind, but not for bike motion. In still air, instantaneously, wouldn't the air see the wheel spinning around the contact patch?

If it was rotating around the hub, that would make it look like it's going backwards at the contact patch.

tanhalt

I think you guys are looking for something like Fig. 1 in this article...

tanhalt

That might be possible...when are you going? It would be nice if the wheels were tested alone, but I'd settle for a comparison on the bike to a C2 Jet 9 with a 23c tire on it.


Don't get blinded by the Crr value. Depending on your race speed, Crr becomes much less of an issue and better aero performance can make up for a lot of Crr "sins". See the plot below which is based on a combined aero drag + Crr value for various tires on a 50mm deep Bontrager ACC wheel. The aero drag was an average from 0-15 degrees IIRC. As you can see, in this case, with apparent wind speeds at anything over ~17 mph and the RXL Aero TT 19C tires (the lighter versions - weigh ~180g) are the "kick ass" combination

Brian Ratliff

Depends what kind of answer you want. You need a 3D rotating solid body and a good turbulence model to start with to even get close to single digit uncertainty for drag. It's a pretty tall order for both the engineer setting up the model and the computer computing the solution.

With a PC, you can get a pretty picture, and that's pretty much it on a complex flow such as this.

Brian Ratliff

I can guarentee you wind does not hit the trailing edge of the rim in such a clean manner. I think the sailors call it "dirty air" and it comes from the leading edge, spokes, hub, and fork. Plus the spokes are rotating and causing turbulence all their own.

Brian Ratliff

Bike motion can be fully reduced to wind speed. Like sitting the bike on rollers in a wind tunnel. On a windless day, the rotation speed of the wheel simply equals the air velocity so the airspeed at the contact patch is net-zero, and the airspeed at the upper part of the rim is net-double. And thinking about it, in a tailwind, the airspeed at the contact patch could be net-negative if you were going slow enough.

That brings up a good question though... since we are looking at 1-2% resolution in the wind tunnel, and spoke rotation causes a fair amount of drag, I wonder if wind tunnel tests ever keep wind speed constant and measure drag with varying wheel speed.

wens

And, having played with CFD some, I can say that you're going to need a relatively large amount of computing time to even get that, much less compare several different options and iterations of those options.

Brian Ratliff

Yup. One of those "hit 'Enter' on your way out the door at night and come back the next morning to find your solution didn't converge" types of things .

tanhalt

It doesn't take much yaw for the back half of the wheel to be out of the wind shadow of the front of the wheel. Take a look at this vid by Zipp's Josh Poertner taken at Interbike. He shows a CFD plot that's a horizontal section with the wind at 15 deg yaw.

https://www.youtube.com/watch?v=CDvi7Ne49qY